Alex Nowrasteh
Politically fueled panics rise and fall in the United States with amazing rapidity. Something rare happens, and everyone pretends it’s common—a terrorist attack, a specific kind of crime, a freak death, or inappropriate behavior by a teacher, politician, or police officer. Sometimes, there are even copycat crimes that seem to reinforce fear. The media will often report on different tragedies, implying that they are somehow similar. After a few days or weeks of panic, it dies down. After a time of relative quiet, a new panic rises. One of the more recent fears was the supposed rising prevalence of politically motivated killings.
The assassination of conservative activist Charlie Kirk was one of those events. Tyler Robinson was motivated by his progressive beliefs to murder Kirk on September 10, 2025. The video of the assassination is shocking, and Kirk was a well-known fixture on college campuses and in conservative circles. His face was especially recognizable to the young and terminally online.
Shortly afterward, the airwaves and social media were filled with stories about how common these types of killings are and how left-wingers are often killing right-wingers. The Trump administration responded by shifting the domestic counter-terrorism state toward targeting left-wingers. Normally sober people had inane comments like, “This time it just feels different.” The facts disagree, as I argued shortly after Kirk’s murder. Today, we just released an improved version of that research titled “Politically Motivated Killers: 51 Years of Terrorist Murders on US Soil, 1975–2025.”
Politically motivated killings are rare. Of the over one million murders committed on US soil since 1975, a mere 3,577 were committed by killers with political motivations. Over 88 percent of those murdered were committed in the two largest attacks that are statistical outliers: 9/11 and the Oklahoma City bombing, committed by Islamists and right-wingers, respectively. The remaining 430 deaths committed by 186 killers who were motivated by political ideology represent the extent of the threat over the last 51 years.
The annual chance of being murdered in a politically motivated attack is about one in 33.1 million since 1975. In more recent years, it’s about one in 19 million. That is a tiny risk, roughly equal to the chance of being killed in a lightning strike in the UK. Every murder is a tragedy, and some have more serious political consequences than others. But that doesn’t justify exaggerating the danger. There isn’t even a statistically significant increase in the number of victims over time, as the deadliest triennial period was 2014–2016 and the safest was 2005–2007.
People are most interested in the ideological breakdown of the killings. With all outliers excluded, right-wing politically motivated killers are responsible for 195 deaths, or 45 percent of all killings, followed by Islamists at 32 percent with 139 deaths, and left-wingers bringing up a distant third place with 68 murder victims at just under 16 percent. With the outliers removed, that’s about 2.9 people murdered by right terrorists for each person murdered by left-wingers in attacks. Some critics strenuously objected to removing outliers. Keeping them in, the ratio of right-wing deaths to left-wing deaths rises to 5.3, meaning that 5.3 people are murdered in right-wing terrorist attacks for every person murdered in a left-wing terrorist attack. Islamism also becomes the deadliest ideology by a long shot with outliers included.
My original blog post on this topic, which I posted after Kirk’s murder, identified 3,597 people who have been murdered in politically motivated attacks. Many critics sent me objections, and some identified killers whom I should not have included. Their hard work convinced me to reduce the number of victims by 0.56 percent to 3,577.
I used an intercoder reliability test to make sure I wasn’t biased in assigning ideological motivations to the killers. I gave three other researchers my ideological definitions for Islamism, right-wing, left-wing, foreign nationalism, separatism, and unknown/other, as well as the list of killers. They then independently researched and coded the ideologies and agreed with my assignment of ideological motivations 95.7 percent of the time. The raw percentage chance of agreement isn’t good enough because it doesn’t control for chance alignment. Thus, we use Cohen’s kappa as a chance-corrected measure. The estimated result indicated near-perfect agreement and is consistent with Krippendorff’s alpha, another measure that is robust for missing data and uneven category distributions. Scott’s pi, another method that relies on pooled category proportions rather than coder-specific margins, returned an almost identical estimate. The convergence of these different statistics is evidence that coder agreement is not an artifact of the underlying distribution of ideological categories.
The remaining coding disagreements were mostly limited to boundary cases in which the coders classified foreign nationalism as right or, in a few instances, as unknown/other. No category was large enough to inflate agreement, and the coders were in perfect agreement for rare categories such as separatism (n = 3). Even if every disputed case were recoded against my judgment, the ideological distribution of the attackers would not meaningfully change, though there would be a few more victims of right-wing attacks.
My paper also includes explanations for my most controversial inclusions and exclusions, such as listing Rahmanullah Lakanwal, who murdered a national guard member in November 2025, as an Islamist even though there is little evidence that he was ideologically motivated and he hasn’t been charged with a terrorist offense. Contrary to what you’ve heard, very few people support committing violence for political ideologies. Bad polling data is ubiquitous on this topic.
The best criticism of my work here is that some victims are more politically significant than others and that their murder can cause serious problems, whereas a greater number of victims in politically motivated mass shootings has less of an effect. The 1960s assassinations of President John F. Kennedy, Rev. Martin Luther King Jr., and Sen. Robert F. Kennedy Sr. were just three murders in a turbulent decade, but they shaped political trends, fomented widespread riots, and prompted other forms of political violence that affected millions. Thus, statistically weighting the murders of prominent people such as Charlie Kirk more heavily than the victims of a mass shooting is intuitively appealing when judging the intensity of politically motivated violence. If there were a consistent and transparent means to weight those lives more heavily when judging the political effects of violence, then it would be worth employing that method. But such a means is absent here, and I don’t know of a good way to do so that isn’t just arbitrary.
Below is a list of all the politically motivated killers.
https://datawrapper.dwcdn.net/6Aoow/1/